The complexities of smartification: Exploring horizontal tensions in smart city governance

Governance in Swedish municipalities

Swedish government structure comprises three levels, the smallest of which are the municipalities, containing both urban and rural areas. Municipalities are responsible for a variety of functions ranging from operating public schools, kindergartens and elderly care (their ‘core tasks’) to planning and managing local infrastructures such as streetlights, water, and sewage systems, housing, and public buildings, sometimes including harbours and airports. Municipalities hold a so-called ‘planning monopoly’ in Swedish law (Plan-och bygglag, 2010: 900 1. kap. §2), meaning they oversee spatial planning at all levels within their territories, except larger infrastructural networks and projects of national interest, e.g. the national roads and railway networks, and certain airports.

Previously most infrastructure was operated by the municipalities themselves. However, since the 1980s there has been a move towards marketisation where public organisations operate following private sector logic while retaining their original goals of serving the public good (Erlingsson et al., 2015). This has meant that many parts of infrastructures have been separated from the municipalities and turned into municipally owned corporations (MOCs), which could be characterised as a form of hybrid governance. The number of MOCs has tripled since the 1970s and today there exist some 1800 MOCs in Sweden, one-fourth of which operate within the water and energy sectors (Erlingsson et al., 2015). In the municipality under study, the airport, the harbour and the water-, sewage- and waste-management, among others, are MOCs. While the water-, sewage- and waste-management corporations operate in a monopoly position, the airport corporation competes in a market with other municipal airports in the surrounding area.

While MOCs to some extent have been studied in the context of urban governance in Europe (Leixnering et al., 2021; Wansleben and Neumann, 2024) and the dangers associated with this form of hybrid governance have received attention (Erlingsson et al., 2015, 2018), they have rarely been studied in the role they play in the development of smart cities and the tensions that arise between them, their municipal owners and other levels of government. Through this article, we hope to start closing this gap.

In the context of smart cities, governance has recently been the focus of several studies. Taylor Buck and While (2017) have pointed to the governance difficulties in turning a smart city imaginary into meaningful urban projects and the tensions that can arise between different levels of government with diverging priorities. McGuirk et al. (2021) explore the ways municipalities have extended entrepreneurial statecraft repertoires to enact smart cities. These include both what they call ‘extrospective’ or beyond-the-state repertoires such as public–private partnerships, sponsoring and how experimentation around smartification is organised and building legitimacy around smart initiatives, as well as the more ‘introspective’ repertories that are directed inwards into the municipality itself. These practices include reconfiguring and reorienting municipal institutions, collaboration between municipalities and developing agile capacity, partly to circumvent traditional procurement processes deemed not fitting smart technology development (McGuirk et al., 2021). Similar to our argument here, scholars have argued for a more complex view of how smart city policies are at the same time glossy as well as fragmented and partially random (Wathne and Haarstad, 2020) and how smart cities are governed beyond a corporate-led agenda (Dowling et al., 2021).

How smart cities are made is thus composed of a range of practices that interweave and might even contradict each other. These include the planned activities that municipalities engage with an eye to economic redevelopment and rebranding, as well as the more after-the-fact activities that can be tagged on to the smart city agenda while they are happening or be integrated into the story of the municipality’s smartification process. In response to Dowling et al. (2021: 3300) assertion that ‘how, opportunistic and purposeful smart city processes intersect in specific contexts’ is not well understood, we here offer a more granular picture of how smart cities are developed, focusing on the practices of the actors involved in doing this work. In so doing, we are also in conversation with the scholarship on smart cities and complexity which often highlights how the deployment of so-called smart technologies produces a simplified picture of the urban space.

The example of the specific responsibilities of Swedish municipalities hints at the spatial and temporal flavour of the complexities in the context of this project, with responsibilities for different areas and tasks distributed and redistributed during the last few years. When we approached our material, these were aspects that we wanted to reflect on in our analysis, without simplifying the tangled picture of actors, technologies, discourses and policies. How then might we put complexity into action in our analysis of a smart city project?

Complexity and smart cities

Scholarly and popular literature has long acknowledged the complexity of the city, grappling with how to represent the many levels, scales and actors that meet in such spaces. The emergence of the so-called ‘smart’ city is the latest variant of this narrative and has been given particular urgency by predictions that an ever-larger proportion of the world population will live in cities ² and the pressing need for sustainable solutions. As noted earlier, much of the literature has drawn attention to how the complexity of the urban environment may be simplified in the implementation of so-called smart technologies. This can be through techno-optimistic views on technology’s neutrality and what technology can achieve (D’Ignazio and Klein, 2020) or through data-driven processes taking the upper hand over traditional planning practices (Kitchin, 2017; Kitchin et al., 2015).

Smart technologies reconfigure how people engage with their environments and cities in multiple ways (Gabrys, 2014; Rose, 2020; Zandbergen and Uitermark, 2020) and thus actively produce urban space (Enlund et al., 2022; Kitchin et al., 2015). When smart technology is poorly implemented it can thus increase social divides (Cardullo and Kitchin, 2019; Harrison, 2017; Robinson and Franklin, 2021). Juvenile Ehwi et al. (2022) find that while issues of data have tended to be the focal point of ethical considerations in smart cities, city officials involved in a smart city project were aware of the legal issues that govern technological implementation. However, ethical and governance principles are less at the forefront of their decision-making (Juvenile Ehwi et al., 2022). Smart technologies are thus embedded in power structures and fraught with representational problems (Burns and Wark, 2020; Cockburn, 1985; Hacker, 2017; Wyatt et al., 2000) and are one of a broader range of tools used in the entrepreneurial governance of cities (e.g. Harvey, 1989; Hollands, 2008; Lauermann, 2018; Peck, 2014). As such, it is not uncommon in these accounts to see ‘top-down’ stakeholders represented as being the major proponents of smartification.

In an effort to manage the inherent complexities of the city space, implementation of smart solutions can often produce simplified understandings of how citizens actually use the space. These simplifications can take different forms. It is pressing that we develop ways to grasp (even if temporarily or provisionally) this complexity to engage with a more holistic view of the emerging urban space. The case of Norrköping illustrates a number of the issues discussed, exemplifying both hybrid governance and fragmented/messy attempts to smartify.

Complexity as an analytical lens

Whilst we see our work as being indebted to the literature on smart cities and complexity in terms of empirical focus, here we want to draw on complexity primarily as an analytical tool to help us. As noted above, complexity is a theme that echoes through much of the scholarly debate around smart cities at present in attempts to develop both theoretical and methodological approaches that can handle multiple priorities. The search for ways to understand, analyse and represent the specific complexities of a place poses a challenge for researchers, and has been connected to the challenge of how to connect localised case studies with global comparative research (Miller et al., 2021). The challenge for researchers is thus two-fold. Firstly, how to capture and represent a complex knot of localised interests and secondly, how then to present it in such a way that it becomes a useful comparator for other scholars engaged in smart city research. The metaphor of the knot is helpful here in that it acknowledges how ‘smart city programmes are always bound up with bits of other projects, processes and agendas’ (Miller et al., 2021: 668).

This interconnectedness poses an additional challenge to those researching the ‘smart city’ as drawing boundary lines around the research study or ‘object’ can be methodologically tricky as well as politically charged. What techniques then can we adopt to help us get comfortable with this? In our introduction we described complexity as an ‘analytical lens’ that we would use to explore ‘horizontal tensions’ around smartification. An analytical lens is usually deployed on material to parse the material, to pull out themes from material or to identify tensions or resonances across materials in order to draw broader conclusions. Understood in this way, an analytical lens can often simplify or crystallise the material into two or three clear ‘results’ which may be presented. In our research, however, we wanted an analytical approach that somehow ‘stays with the trouble’ (Haraway, 2016) of complexity, and which would allow for the results being messy or ambivalent. This is perhaps particularly challenging to attempt in the limited space of a journal article but we feel it is the one that does most justice to what we found in the material. In order to achieve this, we approached complexity not only as a distinguishing empirical feature of the smart city, but also as an analytical lens.

The turn to complexity is now several decades old and has been comprehensively theorised and discussed in the social sciences and elsewhere. Urry (2005) dates it back to work published in the late 1980s and early 1990s, particularly within the natural sciences, with the social sciences catching up in the late 1990s. Urry writes ‘Complexity though is not the same as simply complicated. Complex systems analyses investigate the very many systems that have the ability to adapt and co-evolve as they organize through time’ (Urry, 2005: 3). There is a liveliness to Urry’s definition that fits well with thinking about urban planning, with many forces pulling against or together and a corresponding ripple effect of consequences with each change. Byrne (2005: 96) focuses on methods, arguing that ‘complexity must become more than a metaphorical apparatus in social science and this can only happen if the complexity frame of reference shapes the actual tools of investigative social science themselves’ (see also Cilliers, 1998).

To achieve this we turned to John Law and Annemarie Mol who pose two key questions which are useful here: ‘What is complexity and how might it be attended to?’ (Law and Mol, 2002: 1). Here the contributors share accounts of very different contexts and artefacts, but each is attuned to the inherent complexity of the development/deployment process. In so doing, they provide useful examples of how one might practically represent and understand the convergence of multiple perspectives, agendas or actors involved in an artefact whilst acknowledging the contexts in which each example is embedded. This is an approach that explicitly recognises multiplicity and thus challenges neat boundary lines between different methodological or epistemological approaches. Law and Mol (2002) describe this approach to complexity as:

Imagine looking at different pictures, one after the other. Each orders and simplifies some part of the world, in one way or another, but what is drawn is always provisional and waits for the next picture, which draws things differently. (Law and Mol, 2002: 7)

Inspired by this, and in an effort to produce a more open-ended, ‘provisional’ picture faithful to the complexity involved in smart cities, we decided to use a series of vignettes developed from interviews as a way to handle the complexities experienced in this research. This gave us a way to represent the mixture of discourses, technologies, politics and people we encountered in these interviews.

The vignettes are designed to provide ‘different pictures, one after the other’ (Law and Mol, 2002: 7) of Norrköping. They are based on interview transcripts and convey the voices of some of our participants, putting different perspectives into conversation with one another in an effort to reflect what we might call local complexities. We have suggested some themes around which the vignettes are organised, but as will become clear the themes are somewhat unruly and often return in other vignettes. Rather than try to ‘tidy up’ these results, we view this as part of the sprawling, boundary-less complexity which emerges even in a medium-size city. In this way, we are trying to produce a more realistic account of smartification as it is taking place here and now. Thus, we turn to introducing our methods and material.

Vignette 1: Time

Interviewee 9 described how their MOC is facing a technological shift within the waste management sector which means they have tried to find procedures that would allow them to experiment more with new technologies. ‘I don’t work with daily deliveries today, but I do work with deliveries tomorrow and in ten years, and in what ways we should develop in order to be relevant to the customer’, says interviewee 9 when describing how they approach this ongoing technological change. Right now, they are in the exploratory phase ³ in this shift towards needs-based waste management and electrified vehicles. He continues:

I see before me that in ten years I cannot believe that we will be driving around all town and empty waste containers just because it is Thursday week 22, but rather it will be when the customer wants it to be emptied.

Entering this technological shift has consequences for the corporation that needs to deal with change taking place at different time scales. For example, some of the waste management has been brought back in-house and is once again operated by the MOC itself (rather than the third-party contractors who currently handle daily waste management) to be able to experiment with new technologies. This is due to the scheduling of public procurement processes. Normally, these take around one year to conduct and the contract runs for two years. However, in the case of waste management, the investments are high for the contractors putting in a bid, and to get serious tenders, the contract needs to run for a longer time, maybe up to 8–10 years, to make the investments worthwhile for the contractors. But for the MOC to be bound by a contract for such a long time in these times of change is potentially problematic. This would make experimentation more difficult as contracts need to be renegotiated between the MOC and the contractor operating the waste management for each new experiment, which is further complicated by their divergent interests. It is in the interest of the MOC to improve services in the long-term, simultaneously making them more efficient and climate-friendly. However, for the contractor operating the actual waste management, technological advances that result in driving fewer routes mean earning less today.

Vignette 2: The role of the municipality

Interviewee 2 works at the municipality with issues of long-term development of the municipality through the investigations that this person conducts. These investigations can run across a wide range of tasks, from kindergarten, schooling and elderly care to land use, land exploitation, urban infrastructure development, and building permits, that municipalities oversee and operate. When talking about the job the person highlights that the focus on managing the everyday ‘core tasks’ can block the longer developmental horizon from view:

A lot is about issues of development, and development issues are fun but also difficult. Because … you cannot always expect that the ideas you have, that others will be interested or that you will reach through the noise. Many of the municipality’s issues are those that one does on a daily basis and those that are here and now in the core tasks. It can therefore be difficult to catch the attention [of politicians and colleagues] with future-oriented and developmental efforts.

At the same time, if there is no will from the politicians it makes it difficult to develop an issue, as interviewee 2 concludes:

You can write up a case for decision by the municipal board, but if it doesn’t have a bearing in some type of wish from politics, then it is likely that not much will come from it, then it is hard to reach out with those issues, and the organisation in a way becomes rather big and impenetrable.

Another difficulty balancing the daily operations with future development is imposed by budget restrictions. Interviewee 2 remembers that a few years ago there were suggestions to cut the Education Office Research and Development budget, which was seen as preferable to cutting among the core educational tasks. For the municipal employees, it can thus be difficult to balance the roles of carrying out the politics set forth by the elected politicians, while at the same time being the experts that should deliver investigations and advice based on their expertise rather than politics to the politicians. This tension becomes visible in the prioritisation between long-term developments such as smartification and operating the everyday municipal services, especially under conditions of economic scarcity.

Another aspect of the role of the municipality was highlighted by interviewee 10 when talking about the work with their long-term traffic strategy:

We can build as much city as we want and say that it will be serviced by public transportation. But nothing says this will happen if the public transport company doesn’t agree that it is a good idea in this place. So, we must coordinate and talk to each other.

While the municipality works on building the city, the public transport company oversees delivering public transport in the entire region.

Vignette 3: Infrastructure

‘It’s bloody complicated this regional development work’, says the first interviewee as he sketches the history behind the creation of a logistics platform. His account of a regional effort to identify expertise in various sectors and to align these involves local, national and international companies, the EU, the municipality and the local university. Taking the example of more sustainable goods transport, he describes a recent event where a range of actors gathered to discuss the move to using electric goods transport: ‘There are many who will be affected in different ways, even if no one knows exactly how’ he says as he outlines some of the challenges involved. By way of illustration, he explains how ‘an electric truck costs three times as much as a regular truck’ leading to a requirement to drive more to make them affordable, which is in tension with city regulations that determine that deliveries must not be made on certain streets at certain times. Furthermore, the electrification of goods transport will require an enormous investment on part of the Energy Authority to provide the infrastructure. His account puts the focus firmly on human efforts to work in a more coordinated way: ‘we must be smarter and see cities as a whole system’. Similarly, the eighth interviewee who works for the municipal harbour corporation, resists the tech-centric rhetoric when saying ‘I think it is smart – in response to your question about what a smart city is – when the business is well balanced’. In his vision, balance and coordination between different businesses are what makes a city ‘smart’, not the development or deployment of innovative technologies per se. For those who see the bigger picture such as these two interviewees, there is a high degree of pragmatism and a reliance on human cooperation to produce cities that are good places to live. Their accounts ground the promise of smart cities in logistical and economic realities, as well as good working relations, that highlight the complex interplay of stakeholders, agendas and timeframes.

Vignette 1: Time

The first vignette shows the complexity related to the time when different actors operate in a shared system that has moved from where providing a public good is in tension with a focus on regulating a marketplace where a service needs to be provided. Using time as a focal point, we unpack vignette 1 to find several strands that are knotted together in the case of waste management, including money, actors, technology and governance logic. The MOC is willing to invest in longer-term technological innovation, motivated by working for the public good. However, the actual daily work of collecting waste is carried out by smaller actors (subcontracted waste management companies) who are motivated by business survival. Investing in ‘smart’ sensing technologies in order to drive fewer rounds to collect waste, means less income for these actors who do the work of managing waste. In the background are the companies that can provide technological solutions for ‘smarter’ waste management, and who operate on a longer time horizon, who can convince the MOC about the potential benefits of this long-term project. Their governance logic is long-term commercial innovation connected with profit.

This vignette exemplifies recent work on smart city governance (McGuirk et al., 2021) which has highlighted how urban entrepreneurialism and governance do not always neatly align. Examining the example of waste management by focusing on different temporalities allows us to reveal some of these moments of misalignment or ‘horizontal tension’ and connect them to different strategies and priorities. These differences give clues as to why the roll-out of smart solutions is not always smooth or successful, and points to reasons why some of our participants worked hard to resist the term ‘smart’. By focusing on time, it enables us to bring together future-oriented technology providers with those doing the everyday work of maintaining a city. This connects smart city visions to contemporary realities to develop deeper understandings of what determines the successful development and deployment of smart technology.

McGuirk et al. (2021) are among scholars advocating for closer attention to be paid to the role of the state in smart city initiatives and to avoid mistaking techno-solutions for state governance. In McGuirk’s analysis they focus on two cities in Australia, highlighting the challenges for major urban areas. This focus on larger urban areas is not uncommon in the literature, but may limit the relevance of the findings for smaller cities such as Norrköping. With our focus on a smaller city in a very different context, we are able to reveal sociocultural specificities in ideas about governance as well as what happens in the gap between techno-hype that pushes smart solutions as the answer and the mundane limitations of a not particularly affluent municipality (Taylor Buck and While, 2017).

Vignette 2: Role of the municipality

In this vignette, we explore the role of the municipality as a nodal point, identifying the following strands: time, space, actors, and agendas. The second vignette highlights the different politics and logic that guide the different municipal actors in how to prioritise between providing the day-to-day services citizens expect and the longer-term potential of innovating and making service provision more efficient. It highlights the role of future visions for the city, and the question of how these are to be realised in a context where MOCs, politicians and municipal employees must cooperate. There are several actors mentioned in this vignette by the two interviewees, some of whom are more clearly described than others. The public, for example, remains a constant hovering presence without being explicitly referred to. We see their role as the audience for the politicians and those directly impacted by the services provided by the municipality, concerned by both the day-to-day realities of life in the city but also with an eye on the longer-term development. The strategists at the municipality can afford a longer-term vision of planning for how the city could look in 20 years (with presumably the citizens’ assumed needs in mind), while still needing to accommodate the politicians’ wills. The municipal employees must therefore juggle both competing time frames (everyday versus 20-year view) and different audiences (individual politicians versus an assumed public). This interviewee’s view therefore places the current set of smart city initiatives into a bigger picture of ongoing urban development work that takes place regardless of political shifts or the latest trends. It functions as a salient reminder of the ongoing needs of the population and echoes the point made by Taylor Buck and While that ‘the smart city discourse (including its critiques) is often rooted in the expectation of transformational systemic change that overlooks the roll out of the smart city through multiple incremental and smaller scale changes’ (Taylor Buck and While, 2017: 504). This interviewee’s reflections also raise questions about the ownership of smart initiatives when introduced into a context characterised by tensions between long-term municipal employees and time-limited political positions.

Whilst the municipal developer highlights time in relation to politics, the second interviewee’s reflections bring space more clearly into the discussion. The mention of the public transport company also draws attention to the different spatial configurations at work. A pertinent example of this came to our attention after the interviews were completed when one of the authors noticed the rebranding of all public transport in the region with posters that read ‘Varje resa räknas. Välj smart och hållbart’ (‘Every journey counts. Choose smart and sustainable’ (author’s own translation). The Region prides itself on using either electricity from sustainable sources, or biogas, to power its public transport. While our interview material shows ambivalence towards ‘smart’ technologies in the Norrköping municipality, the regional public transport company appears to have a more positive take on the term.

Vignette 3: Infrastructure

The third and final vignette puts the spotlight on the coordination required when attempting to develop and construct new urban infrastructures, coordination that reaches, in a sense, beyond the state (McGuirk et al., 2021) or in our case beyond the municipality. In this vignette, we hear from two actors with responsibility for overarching planning and coordination in the city. Both hold positions in MOCs, which involve connecting different actors and agendas and making connections to develop long-term infrastructure. In this sense, they represent top-down stakeholders who are usually attributed with an entrepreneurial agenda (Harvey, 1989; Peck, 2014) as their roles revolve around trying to partner the public with business or business with business. Nevertheless, our findings contribute to the growing work of complicating the alignment of entrepreneurial agendas and smart cities (McGuirk et al., 2021), as they are not entrepreneurial in the sense of being ready to hand private businesses free reign to implement ‘smart’ technologies, rather they emphasise balance and adaptive governance when technological developments go into unchartered territory, as exemplified by the electrification of goods. Their positions thus allow them to see further afield than the city and consequently, this vignette brings in national and international frameworks, as well as time, money and a variety of ‘interstitial’ actors (Burns and Welker, 2023). What becomes clear from their narrative is the value of coordination and integration of different processes in municipal statecraft, this could involve new technology but is much more likely to depend on different stakeholders communicating.

In particular, the third vignette draws attention to the different governing and governance logics of national/European/global sustainability policies, local quality of life goals and regional logistical benefits (see also Haarstad et al., 2024) and the actors behind them. Burns and Welker (2023) have pointed to the need to look at the various interstitial actors who influence how smart cities manifest. While we agree, we want to highlight the importance of the negotiations of ethical aims and governance logic that happen in what would often be characterised as top-down actors, but among which we argue are a range of interstitial actors. In ongoing smart city developments, the ‘top-down’ actors have a broad array of other interstitial and ‘top-down’ actors with whom they need to negotiate and find common ground to achieve their goals. These involve different actors ranging from the municipality, private tech companies and MOCs, to more clearly top-down actors such as national agencies and the EU, each of whom answer to different people. While the EU, the municipality and the national agencies answer to the public (often in terms of local people), they can also answer to society in terms of providing the greater good for most people. However, MOCs like the harbour answer to the municipality and businesses, local, regional and global, in terms of providing an infrastructure for flows of goods and connecting diverse businesses that might benefit from synergies in their businesses due to opportunities for collaboration. Our findings are thus in line with those of Dowling et al. (2021), highlighting how making smart cities is ‘simultaneously contingent and piecemeal and purposeful and strategic’ (p. 3311).